Archicad Future, 27 Poll

[bouhmidage]

Archicad 26 was released yesterday, personnally i'm disappointed of how Archicad is slowly developped in the last 4 years, the pandemic has it' effects on business for sure, but we expected more since earlier versions used to bring nice and cool features, 

I created this tread for users who have ideas, suggestions, propositions for the next release, we may express our thoughts, and give guidelines for development team, Graphisoft is listening, i'm sure, 

For me, the most important thing to in an urgent way is to stop this multilingual installers strategy, 

Archicad 26 is released, all the world can download it with a multilangual installer,  : 

1 - download and install in a preferred langue, 

2 - open Archicad using your template, or download a template from the website, almost of us migrate tempaltes from older versions, resellers template  are useful for new users, 

3 - libraries also can be dowloaded from website, this will help in 2 things : lignter installers, and libraries are accessible for everyone especially when user works for foreign companies, .

this way, development and bug fixes will be focused on 1 installer, not 10, 

----------------------------

Archicad strategy is to implement several disciplines, let's admit this, 

For users , Architects, engeneers, drafters, commands and disciplines should be seperated and organized in different way, so we don't feel abused with unecessary command, 

yes Archicad offer possibility to organise commands in the work environnment, but the saved commands layout will be a mess when upgrading to newt version wich contain new commands, 

why not,  from a buttons menu you can activate ad desactivate disciplines commands.

----------------------------

You said in the webinar, Focus on design, Archicad tools are really good and that's why we stick to Archicad , BUT aren't fully baked, workarounds and illogic solutions are always there to disturb the workflow, think about new users, workarounds make them feel uncomftable, saying, " this simple situation needs a workaround, whatabout complex ones ?? " 

each tool have major / minor things to update, tweak , to get a good design toolbox for daily work, 

to attract users, Archicad should easily be able to help designing a residentiel or midscale projects without workarounds, this kind of project is where users start exploring the software, if it fails, users will start searching around for alternatives, 

make the product a complete solution for architects, 

----------------------------

i think , Archicad 27 should be the " NO workarounds version "   ,make architects life easier, to focus on design , then you can go ahead for structural and MEP workflows, 

[David Shorter]

Hi Jim

From your comments I guess you're in the UK.

I'm not sure I understand "You missed the point about instanced elements. This is potentially a game-changer in terms of productivity." Can you clarify pls? Your comments about the Mesh are interesting but need more of a discussion. You probably don't remember ArchiSite, now that was a disaster.

Can you send me a screenshot of your Accessories library I suspect I have more here ( inAustralia )

And yes a shared doc is what I had in mind.

And and it may be a waste of time but doing nothing will achieve nothing

[jl_lt]

Another thing that might help in all this endeavour while posting wishes is this: trying to diferentiate between desireability and our own desire;  So asking ourselves "this thing i want will actually be helpful for others in all kind of projects?"  Might be a useful mindframe to begin. Graphisoft should do this too 

[Jim Allen]

Agreed, that's why I've added a colum to the sheet for justification. Sometimes we don't think we might  something, but when it's explained in detail, we might think differently.

[Jim Allen]

With instances, the idea is that it mirrors an actual object's configuration. 

There are lots of ways you can use this, and placeholders is a particularly good example.

Say you have a large office layout and you want to add furniture to it. You can define a table type, and everything will be an exact duplicate of that type - an 'instance' of it. So you can change the manufacturer, finish, colour, dimensions etc, and any change propagates to all the duplicates.

Doors are a better example. In any given project I will have say 6-maybe 20 types of doors. I define doors as having ironmongery packs, so you can have say just 6 packs to cover a whole building. 

Doors have different widths, vision panels, signs, fire resistance, finishes specification clauses and of course ironmongery.

Every StoreDoor1 could be defined and updated just once, with all the changes automatically propagated throughout the model. If I change a specification clause and modify a couple of door types to change the parameters that reference the specification clause, everything will be updated once the master definition is.

I started working on a model that had some multi-segment column in it. One segment for the flange plate, a middle segment for the column itself, and an upper segment for the fixing plate. With an instance I could start with a simple column and refine the detail once with every instance updated. How much time would that save?

Say for example I wanted to change the hand driers in a model, I could change the master instance and it's done very, very quickly.

In effect it's like Vectorworks symbols, Autocad clocks and SketchUp components.

I can post screenshots of my library, but I'm on leave right now (hence all the posts) and back in on Tuesday.

[Barry Kelly]

Aren't instances just hotlink modules?

So your column example.

I place a simple column and save it as a module.

I then place that module where ever else I want it (even in different files).

Now, I modify the original column and re-save as the module or better still set up the publisher to do it for me.

The original column you save as the module doesn't even need to be an actual column in the project.

It can be on a separate storey (or even a separate file), specifically for creating modules.

This way it is easier to keep track of the original source.

Now all I have to do is update the hotlinks.

Your example of the doors - different sizes with the same options, is not really an instance as they are not all exactly the same.

That is more 'styles' which you can co-ordinate with schedules.

Create a schedule that you can sort by your door style (StoreDoor1, StoreDoor2, etc.) and then have fields for all of the object parameters you want to control - door leaf, handle, style, hinge type, etc.

Now you can amend the parameter settings in the schedule for each door style, and all of those doors will change and still keep all of their separate sizes.

Or maybe I am not understanding what you are trying to explain.

Barry.

[David Shorter]

from a Graphisoft point of view the question is...

"Will this proposed feature sell more Archicad"

[David Shorter]

@Barry Kelly Both work, I tested schedules which for me are firstly a management tool, modules also work. The issue is that different softwares do things differently, whilst some methods have advantages in some directions others have advantages in the opposite. I don't think I have ever needed to change ALL the placed elements of type but if I did then I'd select all doorTYpe1 in the 3D windows and change there.

I guess I've been conditioned to Archicad for far too long 😀

Perhaps we need to analyse  the various workflows

[Jim Allen]

Probably right, but arguably what they also ought to be asking is, "Will these features help to promote Archicad and retain existing customers?"

[Jim Allen]

Hotlinks are a bit of a mess for me. First of all they don't work with Solo, which ought to be essential.

Secondly, they are not proper building objects. A hotlinked door doesn't behave like a door in that it can't be inserted into a wall and scheduled - or can it?

Managing them is messy because they live outside the model, which is not at all what is required - it's too complicated, and it's not what hotlinks are for.

Doors are (can be) instances can't they? Every Cleaner's store door will be identical apart from the element ID and number. Every lockable office door can be an instance apart from the element ID and number. Same goes for toilet doors.

In effect what should work is an extension of the Favourite functionality. Once you define a favourite, it would be useful to be able to modify it, and have an opportunity to select all identical insertions, and reset the properties to the primary definition.

Perhaps we should define the wishlist item like this instead...

[DGSketcher]

In AC doors aren't instances and can't be, as the program structure only allows their placement in context with a wall, they cannot exist in isolation. A wall with a door in it could be an instance. The management of door types lies as @Barry Kelly suggests either with Schedules or maintenance of Favourites. As noted by @David Shorter, you can also Select all the doors tagged / IDed as Type 1 in 3D and apply the Favourites settings to update. It isn't actually a great solution as there are problems with the whole transfer settings process which can be unreliable. I would like to see a robust review of Favourites to improve that reliability.

Hotlinks still have their place in AC, particularly if the attribute divergence issue is resolved. Instancing on the other hand should basically form a robust version of grouping, with the ability to generate clones and isolate the "group" for editing. They should be quick and easy to update without all the mental gymnastics required to track down the source and updating external files. I have never understood GS's reluctance to engage with this issue, other than they are overthinking it. The strangest thing is, if an external .MOD is deleted AC still retains all the Hotlink data, so the database for storing instances is already present in the PLN file structure, and editing is basically as case of ghosting or hiding the elements NOT contained in the Instance, so the remaining elements can be worked on in the same AC session. They can do it for the Curtain wall tool, why not Instancing groups?